Morgan Stanley: "the hype around generative AI may be justified and..."

[Renew Deal]

Link to tweet

Either figure out how to use it or get left behind.

[highplainsdem]

1. Morgan Stanley: ChatGPT will keep 'hallucinating' wrong answers for years to come

https://markets.businessinsider.com/news/stocks/chatgpt-ai-mistakes-hallucinates-wrong-answers-edge-computing-morgan-stanley-2023-2

ChatGPT will continue to "hallucinate" wrong answers occasionally for years to come and won't take off until it's on your cellphone, according to Morgan Stanley.

In a note dated Wednesday, the US investment bank highlighted the AI chatbot's shortcomings, saying it occasionally makes up facts. "When we talk of high-accuracy task, it is worth mentioning that ChatGPT sometimes hallucinates and can generate answers that are seemingly convincing, but are actually wrong," Morgan Stanley analysts led by Shawn Kim wrote.

"At this stage, the best practice is for highly educated users to spot the mistakes and use Generative AI applications as an augmentation to existing labor rather than substitution," they added.

ChatGPT, developed by OpenAI, recently shot to fame after Microsoft injected $10 billion into the company. While its debut kicked off a sudden frenzy in AI stocks, it's also been met with judgement. Academics have warned that platforms like ChatGPT could print misinformation. For example, Insider's Samantha Delouya asked the language tool to write a news story – and it spat out fake quotes from Jeep-maker Stellantis' CEO Carlos Tavares.

-snip-

In other words, ChatGPT has to be fact-checked as if it's a liar and con artist.

It provides what might be true information but might also be partly or completely nonsense in a very authoritative tone.

I've seen it called a bullshit generator, and that's probably the best description.

Fact-checking is time-consuming. Most people won't do it, or won't do it adequately.

Its mistakes include attributing nonexistent articles to real people and nonexistent products to real companies, wasting their time as well when people who expect ChatGPT to be reliable try to follow up on what the AI told them.

It can give wildly varying responses to the same question at different times.

And Morgan Stanley expects this to continue for years.

As for why ChatGPT scaled so rapidly - it's free (though better versions that charge per month are being offered now that some users are hooked) and it helps students cheat. And it does the same thing for adults by doing writing and research for them that they're too lazy to do for themselves. That's the main draw, and those are the people who are least likely to be able to recognize ChatGPT's mistakes immediately by themselves, and least likely to take the time to fact-check its results.

[hunter]

11. How is ChatGPT any different than television news and opinion?

I gave up on those a long time ago. It was too much work fact-checking all the pernicious lies and nonsense.

On top of that there's the advertising, which is 100% bullshit... Ask your doctor.

If I'm reading something I can go as fast or as slow as my critical thinking skills require. And I can quickly skip over the bullshit.

I skip most posts here on DU because they're obviously derived from the television noise machine. Or almost as bad, tweets.

Television news and opinion is just a fire hose of bullshit directly into your brain.

These "Artificial Intelligences" are not doing anything that humans haven't already done. Humans frequently act as automatons spewing bullshit. As author Theodore Sturgeon observed, ninety percent of everything is crap.

These new technologies won't change a thing. They might make the job a little easier for the crap generators but that market has long been saturated with little room for expansion.

[highplainsdem]

12. I'm sorry you're so cynical. But cynicism never does anything to help with

any problem.

You can notice where there are failures and deceptions without giving up.

ChatGPT and similar AI will make people dumber, less creative - and less able to make a living. It will spread misinformation and disinformation even faster.

It will make a very small percentage of people wealthier.

It's a huge step in the wrong direction, and that should not be dismissed as something normal out of cynicism.

[hunter]

16. Every human being is born an artist and a scientist.

Sad to say that's been beaten out of many of us.

I'm never cynical about art or science. My parents are artists who met in Hollywood and they always had day jobs related to their arts. When my dad retired from his day job with a decent union pension my parents became full time artists.

I grew up in a household of chaotic art. You never knew who you might meet in my childhood home, someone sleeping on the sofa or waiting to use the bathroom, future sports celebrities, actors, mathematicians, and social justice warriors... all arts.

But my parents' day jobs were not their art.

My wife and I live in a home stuffed to the rafters with art, books, and music. That was the environment we raised our children in.

One of our kids is a classically trained poet, an English major, and has been paid for poetry, which is possibly the least lucrative of all the arts unless you turn your poems into pop songs or advertising jingles. It's not their day job.

My wife is an artist and a scientist, got the academic degrees, has been paid for both, but not currently her day job.

My college major was Evolutionary Biology and my minor was English. Both arts, neither especially lucrative. When I met my wife we were both teaching science.

My "day job" arts were mostly lab work, construction, teaching, and hard physical labor. Honestly, the physical labor stuff was my favorite, it gave me plenty of time to think about other stuff, theories of everything, but I can't do that anymore because my body is worn out.

[highplainsdem]

18. Hunter, it's great that your family has valued both science and the arts

so much. And kudos to you and your wife for raising your kids that way.

Valuing physical labor is great, too. My grandfather and one of my uncles were farmers.

I would like to see everyone well educated and valuing the arts, science and physical work.

I believe AI threatens all of that. People in the arts should have at least some chance of making a living from it. Blue collar workers shouldn't have to worry that they'll lose their jobs as soon as corporations can get agile-enough robots controlled by AI.

We do not have a society providing a social safety net if almost no one works.

Nor do we have other outlets to give people a sense of purpose and achievement if work and the arts are handed over to AI. Do you really think we're going to have a future where corporations and the wealthy pay enough taxes for everyone else to live comfortably and idly?

Do you think we won't dumb down the population, turning almost everything over to AI?

And there are more immediate concerns. We should not allow corporations to rip off art, music and writing created by humans so AI can "create."

And we should not tolerate AI chatbots that "hallucinate" and communicate in a convincing, authoritative tone even when what's being communicated is wrong, sometimes dangerously wrong - like the recommendation of eating crushed glass from the scientific AI chatbot Galactica that Meta had to yank after just a few days last November.

[Meowmee]

2. We have been warned it may be used to cheat

On writing assignments. I have noticed some students seem to have a much higher level of writing ability but I have no clue if they’re using it or not. Supposedly there is a way to check but I do not have time to start dealing with this nonsense. And I would never be backed up if anybody was cheating anyway however they did it. I hope it fails miserably.

[edisdead]

4. it isn't going to fail anymore than the internet failed

It is here to stay. It is part of reality. And AI in many forms will only become more and more part of our lives just the same as it ever was. It is just the next iteration of progress. It is the wheel or making fire many many iterations over. It will change life as we know it but it will not end life. The hype over it is absolutely crazy to me. The fear over it is a little funny to me. The promise of it instills hope in me.

[Meowmee]

5. Disagree

This is not progress. It will not be a part of my life for sure. Except for Checking to see if people are using it to cheat if I figure that out.

[edisdead]

6. I wasn't talking about YOUR life

You may think that you are not going to be affected by it but you will be by everything that surrounds you depending on how long you live. It is “progress” whether you want to define it that way or not. We haven’t always liked the way progress unfolds but thats life. I am sorry but it is true that AI is going to (and has been) be part of machine and computing from here on out. There will be the negative side of that for sure. But the positives of that will far outweigh the problem of “someone cheated on their term paper”. People are fretting about a chatbot right now. A chatbot….

I totally respect your stance and passion for rejecting it. However, that’s like me sitting here in MN saying the snowstorm that just hit the midwest will have no effect because I shoveled my driveway.

[Meowmee]

7. Time will tell

I remember people here telling me self driving cars would take over as well.

[edisdead]

8. they will.

Of course self driving cars are going to be a thing. Will they be a thing right now? Nope. But they are constantly being developed. That also isn’t going to stop.

[Meowmee]

9. Lol

Nice ideas you have

[edisdead]

13. Most new cars are already assisted

Most new cars are already assisted to the point they know what obstacles are on the road. I drove a ford on vacation last weekend that knew to turn on the brights and to dim them when oncoming traffic was present. Is that self driving? Nope but it is automated processes and “awareness” This os standard on many vehicles now or will be in the next couple of years. Driving is not actually that hard of a task especially in situations like freeway driving and especially during heavy slow moving traffic. Where “self driving cars” and AI will start to shine is during those situations that will help to unravel the human error that exacerbates those situations. I drive for a living. And honestly it is a system. And like any system as long as you follow the norms and recognize when the system is not being followed, it is a mundane and not very exciting activity.

[hunter]

15. When I was in college our exams were hand written in blue books.

I don't see any harm in going back to that.

If the writing style of a hand written five paragraph essay in a blue book is substantially different from that same student's term paper then that's a cause for concern.

My own style of writing is such that teachers and mentors always recognized me, even when I was hoping to be anonymous.

I try to maintain some level of anonymity here on DU, to save my wife and kids any potential embarrassment, but that kind of anonymity hasn't always worked for me. I've gotten the dreaded "You wrote that, didn't you..." a few times in my life.

My mom, who is a much better writer than I am, is a language chameleon and she made some money as a ghost writer putting words into the mouths of celebrities they themselves couldn't find. Not generally a bad thing, they bought the words, they own them.

Some sorts of cheating are very dangerous. I don't want a doctor who cheated through med school and I don't want to fly in an airplane or drive across a bridge designed, built, or maintained by any sort of cheater.

[Meowmee]

17. When teaching in person I always did

Hand written exams. But people can still cheat depending on the type of exam. Now I am teaching online for the classes that have written assignments so it’s a very different situation.

You said you won’t don’t want doctors etc. that cheat on their exams but I can assure you that a lot of your doctors even if they didn’t cheat on the exams are terrible anyway. Not that I’m encouraging cheating on exams etc. But I do not get paid enough and I am not sure how to check for this type of cheating. One of the exams I have I made it fool proof for cheating it’s a multiple choice exam which is graded by a system. All of the questions are shuffled and they come from a pool of 200 possible questions each from a different subject group. If a student has to retake the exam for any reason they will never get the same exam again. The answers to the questions are also shuffled as well.

There’s a time limit however I do let them use study materials so it’s sort of like an open book exam. I can guarantee you though even with me letting them do that there will still be people that try to cheat and they’ll probably figure out a way to do it. And some people still do very poorly on the exam even with being allowed to look at study materials.

[Wounded Bear]

3. Maybe we need to enact the Dune Commandment...

"Thou shalt not make a machine in the image of a human brain."

Something like that. Not that anybody would abey it as a commandment.

[Prairie_Seagull]

10. There will be unintended negative consequences, hell there already are.

Shouldn't that be it's own warning? Shouldn't it be put under a microscope and discussed with the rest of the world? I've said this before, we are all going to meet this beast.

Profit before People

[edisdead]

14. So a world law?